2418
J. Agric. Food Chem. 2007, 55, 2418−2423
Moderate Ferulate and Diferulate Levels Do Not Impede Maize Cell Wall Degradation by Human Intestinal Microbiota CAROLA FUNK,† ANNETT BRAUNE,‡ JOHN H. GRABBER,§ HANS STEINHART,† AND MIRKO BUNZEL*,† Institute of Biochemistry and Food Chemistry, Department of Food Chemistry, University of Hamburg, Grindelallee 117, 20146 Hamburg, Germany; German Institute of Human Nutrition Potsdam-Rehbru¨cke, Department of Gastrointestinal Microbiology, Arthur-Scheunert-Allee 114-116, 14558 Nuthetal, Germany; and U.S. Dairy Forage Research Center, Agricultural Research Service, U.S. Department of Agriculture, 1925 Linden Drive West, Madison, Wisconsin 53706
The degradation of plant fiber by human gut microbiota could be restricted by xylan substitution and cross-linking by ferulate and diferulates, for example, by hindering the association of enzymes such as xylanases with their substrates. To test the influence of feruloylation on cell wall degradability by human intestinal microbiota, nonlignified primary cell walls from maize cell suspensions, containing various degrees of ferulate substitution and diferulate cross-linking, were incubated in nylon bags in vitro with human fecal microbiota. Degradation rates were determined gravimetrically, and the cell walls were analyzed for carbohydrates, ferulate monomers, dehydrodiferulates, dehydrotriferulates, and other minor phenolic constituents. Shifting cell wall concentrations of total ferulates from 1.5 to 15.8 mg/g and those of diferulates from 0.8 to 2.6 mg/g did not alter the release of carbohydrates or the overall degradation of cell walls. After 24 h of fermentation, the degradation of xylans and pectins exceeded 90%, whereas cellulose remained undegraded. The results indicate that low to moderate levels of ferulates and diferulates do not interfere with hydrolysis of nonlignified cell walls by human gut microbiota. KEYWORDS: Ferulic acid; diferulic acids; triferulic acids; cross-links; arabinoxylans; maize cell walls; fermentation; human fecal microbiota; degradability
INTRODUCTION
According to a widely accepted definition, dietary fiber is the edible part of plants or analogous carbohydrates that are resistant to digestion and absorption in the human small intestine with complete or partial fermentation in the large intestine (1). Consequently, polysaccharides, oligosaccharides, lignin, and associated plant substances are included as dietary fiber constituents. Major constituents of cereal dietary fibers are arabinoxylans, mixed-linked β-glucans, and cellulose. The physiological effects of arabinoxylans are dependent on their physicochemical properties, which are related to their degree of polymerization, the arabinose/xylose ratio, the distribution of side chains, and the degree of cross-linking, especially via hydroxycinnamates. Ferulate is the predominant hydroxycinnamate in cell walls of cereal grains, acylating arabinose side chains at the O-5-position (2). Radical coupling of ferulates yields ferulate dehydrodimers (3) with lesser amounts of ferulate * Author to whom correspondence should be addressed (telephone +4940-42838-4379; fax +49-40-42838-4342; e-mail Mirko.Bunzel@ uni-hamburg.de). † University of Hamburg. ‡ German Institute of Human Nutrition. § U.S. Dairy Forage Research Center.
dehydrotrimers (4-8) and ferulate dehydrotetramers (8), which mediate intra-/intermolecular coupling of arabinoxylans (9, 10). Arabinoxylans may also be coupled to lignin (11, 12) and proteins (13) via a radical mechanism involving ferulates, further bolstering cell wall cross-linking. Although lignification is a major determinant to fiber degradability (14), ferulates and diferulate cross-links are also thought to inhibit polysaccharide degradation. Whereas early studies suggested a direct toxicity of phenolic acids on ruminal microbes, more recent studies proposed that ferulate substitution and especially diferulate cross-linking impede binding of endoxylanases, thus limiting xylan degradation (15-17). For example, studies with nonlignified cell walls from maize cell suspensions revealed that diferulate cross-links restrict the rate and, to a lesser degree, the extent of cell wall and xylan degradation by fungal enzymes (18). Other work with whole and fractionated wheat bran indicates that xylan degradation by endoxylanase was related to the extent of arabinose substitution and diferulate crosslinking (19, 20). These restrictions may be partially overcome by esterases able to cleave diferulate esters. However, feruloyl esterases, including the esterase produced by the intestinal bacterium Lactobacillus acidophilus, are much more efficient in releasing ferulate and its dimers from soluble substrates than
10.1021/jf063109k CCC: $37.00 © 2007 American Chemical Society Published on Web 02/24/2007
Degradation of Maize Cell Walls by Intestinal Microbiota from untreated insoluble cell walls (21-24). The aim of this study was to establish whether ferulate substitution and diferulate cross-linking influence the degradation of nonlignified cell walls and cell wall arabinoxylans by human intestinal microbiota. In this study, we used a cell wall model system in which ferulate substitution and diferulate cross-linking were specifically manipulated without altering other cell wall properties. MATERIALS AND METHODS General. Nylon bags (4.5 × 9.0 cm) used in the fermentation experiments were made from polyamide monofilament fabrics (NITEX, polyamide 6.6, mesh opening ) 10 µm) (Sefar, Freibach, Switzerland). Prior to use, nylon bags were heated for 30 min at 100 °C and stored at 55-60 °C to minimize bacterial contamination. Brain-heart infusion (BHI) broth was from Roth (Karlsruhe, Germany). Sephadex LH-20 was from Pharmacia Biotech (Freiburg, Germany). Chemicals used in chromatographic procedures were of HPLC grade. Synthesis of Primary Maize Cell Walls. Cell walls with different concentrations of ferulates and diferulates were prepared according to the method of Grabber et al. (25, 26). Cell suspensions (350 mL) of maize (Zea mays L. cv. Black Mexican) were grown in 1 L flasks. On the third and ninth days after culture initiation, 2-aminoindan-2phosphonic acid (AIP) was added at 0 or 5.4 mg/L to produce cells with normal or reduced levels of cell wall feruloylation. After 14 days of culture, cells were collected at the early stationary growth phase on a nylon mesh (20 µm pore size) and washed with cold 1,4-piperazinediethanesulfonic acid (PIPES) buffer (10 mM, pH 7.0). Cells were suspended in PIPES buffer and ruptured by two passages (15 min) through a Parr nitrogen bomb maintained at 1500 psi. After cell rupture, cell wall fragments were collected on a nylon mesh (20 µm pore size) and washed sequentially with PIPES buffer, aqueous CaCl2 (50 mM), and water to remove cytoplasmic contaminants. Cell walls from normal and AIP-treated cultures were resuspended in water, and then an aqueous solution of H2O2 (0 or 0.3 mmol per gram of dry weight of cell walls) was added with stirring over a 60 min period to promote diferulate formation via wall-bound peroxidase. After an additional 30 min of stirring, cell walls were collected on glass microfiber filters (3.1 µm retention), washed thoroughly with water and acetone, and then oven-dried at 55 °C. Characterization of Cell Walls. Analysis of neutral carbohydrate components was performed as described previously (27, 28). Briefly, cell walls were pretreated with 12 M H2SO4 at 35 °C for 2 h. Following dilution, hydrolysis was performed with 2 M H2SO4 for 1 h at 100 °C. Cellulose crystallinity is a main barrier to its chemical degradation. Therefore, prehydrolysis using 12 M H2SO4 is needed to alter the fine structure of cellulose. In contrast, acid hydrolysis of hemicelluloses using 2 M H2SO4 is possible without pretreatment using 12 M H2SO4. Uronic acids in acid hydrolysates were determined colorimetrically according to the method of Blumenkrantz and Asboe-Hansen (29). Following reduction of neutral monosaccharides with NaBH4, the resulting alditols were acetylated with acetic anhydride by using 1-methylimidazole as a catalyst. Erythritol was used as internal standard compound. Alditolacetates were determined by GC-FID (GC Focus Series, Thermo Electron Corp., Milan, Italy) using a 0.32 mm × 30 m DB-5 capillary column (0.25 µm film thickness) (J&W Scientific Inc., Folsom, CA). GC conditions were as follows: initial column temperature, 150 °C, held for 3 min, ramped at 4 °C/min to 220 °C, held for 2 min, ramped at 10 °C/min to 300 °C, and held for 10 min; injector temperature, 250 °C; split ratio, 1:6; flame ionization detection (detector temperature, 290 °C). He (1.6 mL/min) was used as carrier gas. Phenolic monomers and di-/triferulates were analyzed by HPLC with photodiode array detection (DAD) following saponification and a preseparation using small-scale Sephadex LH-20 chromatography, as will be described in detail elsewhere. In brief, cell walls were saponified for 18 h in the dark using 2 M NaOH (degassed with N2). Monomethylated 5-5-dehydrodiferulic acid and o-coumaric acid were added as internal standards just prior to saponification. The alkaline hydrolysates were acidified (pH
